New tunnel system to eliminate sanitary sewer overflows and control combined sewer overflows in Hartford, Connecticut

The South Hartford Conveyance and Storage Tunnel (SHCST) project is a significant component of Hartford Metropolitan District's (MDC) Long-Term Control Plan (LTCP, 2014), which is overseen by the Connecticut Department of Energy and Environmental Protection. The project will address a portion of MDC's Clean Water Project (CWP), reducing combined sewer overflows (CSOs); eliminating sanitary sewer overflows (SSOs); and reducing the amount of nitrogen released into the Connecticut River (Nasri et al., 2013).

The purpose of SHCST is to eliminate West Hartford and Newington SSOs, as well as Franklin Area CSOs discharging to Wethersfield Cove, and to minimize CSO discharges to the South Branch Park River. The overflow locations are shown in Figure 1.

In 2010, MDC prepared a Preliminary Design Report (PDR) for SHCST, which included relief of the Folly Brook Trunk Sewer and kept the tunnel boring machine (TBM) retrieval shaft within Hartford City limits. Figure 1 shows the recommended tunnel route in the PDR. Since 2010, the objectives of SHCST have changed slightly. The LTCP has been revised and relief for the Folly Brook Trunk Sewer is no longer necessary. MDC has also decided to perform less sewer separation in the Franklin Avenue drainage area. Instead, new relief points are proposed in the Franklin area with diversions to SHCST. Figure 1 shows the current recommended tunnel route (Alignment F).

During 2012, MDC evaluated the potential for connecting the proposed North Tunnel (originally proposed as an independent tunnel with its own pumping station) into SHCST. It was concluded that the two tunnels could reasonably be connected and operated as a single system, utilizing the pumping station at the eastern terminus of the South Hartford Tunnel (Figure 2). It also was concluded that this was less costly than having two, independent tunnel systems.

During dry weather, SHCST will not receive flow as the existing collection system can adequately convey flow to the Hartford Water Pollution Control Facility (HWPCF). During wet weather, when the existing system's capacity is exceeded, SHCST will receive overflows that would have been discharged directly to receiving waters.

This paper summarizes the design basis for SHCST. New diversion structures will be constructed at each CSO/SSO relief point to divert overflows to new consolidation sewers (near surface). These, in-turn, will discharge flow to hydraulic drop shafts, to be conveyed in a controlled manner to the deep storage tunnel. Flows in the tunnel will be pumped to the new headworks at the HWPCF. The components of SHCST described in this paper are

• Deep rock tunnel (6.7 m ID at 6.6 km) with a TBM launch shaft near the HWPCF in Hartford and a retrieval shaft in West Hartford

• 3.7 km of near-surface consolidation sewers (60–170 cm diameter)

• Seven hydraulic drop shafts

• 100 MLD tunnel dewatering pumping station

• Odor control at all potential air release points.

The sizing of the tunnel was based on the volumes from the 1-, 18- and 25- year design storms per the LTCP and updated collection system modeling from MDC's Program Management Consultant (CDM Smith 2014). The LTCP specifies a different level of control for each tributary area. Table 1 shows the peak flows and volumes to be stored in SHCST for each major source and respective design storm.

Surge, air entrainment and pressure waves can occur in CSO tunnels when they fill rapidly, with detrimental results such as geysering, blowback and flow instabilities. The preliminary hydraulic analysis suggests that surge in the SHCST is unlikely, due to the relatively large tunnel diameter in comparison to the incoming peak flows.

Sedimentation can present an ongoing maintenance burden if not controlled. The initial sedimentation analysis and modeling indicate that a 0.1% slope is adequate to minimize sediment deposition issues. This is consistent with the state of practice for other large diameter CSO tunnels as steeper slopes increase project cost. The tunnel will still require periodic maintenance to remove sediment build-up over its life.

Tunnel operation must not result in odor complaints, so odor control has been assumed at each drop shaft location.

An alignment study was conducted to evaluate various tunnel and conduit configurations. Seven (7) conceptual rock tunnel alignments and associated consolidation conduit options were developed, in order to identify a cost-effective tunnel alignment that would balance the expectations of stakeholders.

A systematic approach was established for option scoring. The cost estimate was used as the quantitative assessment for the various options and was not included in the weighted ranking, which is the qualitative assessment. Three stakeholder impact categories were defined, consisting of high, medium, and low impact evaluation factors. Each of these was given a raw score and a weight which depended on its category. The score for each potential alignment was then determined as the weighted sum of all of its evaluation factors. Alignment F was identified as the best option and recommended for final design. It provides the maximum reduction in consolidation conduit length, reducing the associated costs, business impacts and construction risk. Figure 3 shows the configuration of Alignment F.

The site area lies in the Central Lowland physiographic province, which extends north-south through the middle of the state. The area is underlain mainly by sedimentary rocks and igneous basalts of Triassic and Jurassic age. The Hartford Basin of Connecticut and southern Massachusetts is a half graben, 145 km long and approximately 4 km deep, with basaltic lavas and intrusions into the sediments (Hubert et al. 1978). The main source for the sediments was the metamorphics of the Eastern Highlands. Volcanic flows separate the lacustrine and fluvial deposits. Displacements along the faults continued throughout deposition period, and the deposition sequence resulted in various features including alluvial fans and mudflats, lakes, and floodplain deposits.

The last major tectonic episode affecting regional geology occurred during the Jurassic, approximately 175 million years ago (Northeast Nuclear Energy Company 1975). All faults in the project area are therefore considered inactive.

The region has been glaciated many times, the glaciers grinding down local peaks, eroding rock and laying down a heterogeneous layer of till. The till is present over much of the area, overlying the bedrock. The sediments of Glacial Lake Hitchcock filled the deeply incised Connecticut River Valley and lake deposits are present in many places. Glaciers left the area with much of its present day relief. Younger alluvial deposits are common along the Connecticut and Park rivers and their tributaries.

Many different soil types are present in the site area. The formations found in the vicinity of the project and that might be encountered along the route are the Portland Arkose, Hampden Basalt, and East Berlin Formation, comprising shale and basalt with fractured and faulted zones (Figure 4).

The final geotechnical investigation program consists of 55 deep and 50 shallow borings, and 5 geophysical survey lines. The program includes geophysical logging (acoustic televiewer) performed in 21 deep and 5 shallow bores, water pressure (packer) tests in 30 deep and 8 shallow bores, 6 in situ stress determinations in two deep boreholes, falling head tests in the soil profile in selected bores, observation wells installed in 13 each deep and shallow bores, and 22 vibrating wire piezometers installed in 16 bores. The program also included groundwater level monitoring, as well as laboratory tests on soil and rock samples.

The deep rock tunnel – approximately 6.6 km long, with internal diameter 6.7 m – will be excavated by a TBM suitable for hard rock conditions. The tunnel profile is entirely within bedrock deep enough to accommodate the north tunnel system (part of a separate contract). Several types of rock TBMs can operate in various different types of ground condition. They include main beam, single shield, double shield, and convertible (hybrid) hard rock/earth pressure balance (EPB) machines. TBM selection is important as it will impact the type of final lining, construction safety, quality, cost and schedule. The final recommendation on the TBM takes into account factors including rock and groundwater conditions along the alignment, and is based directly on borehole data obtained from the geotechnical investigation program.

It is anticipated that the rock along the alignment will primarily consist of competent shale, sandstone, and basalt, dipping at 10–20°, with occasional fault zones. It may also contain diabase dikes which, if encountered, may contain fractured rock and flowing water.

The size of the construction shafts will depend on the TBM type and diameter, and the dimensions of the permanent structures to be housed in them. For a 7.6 m diameter TBM (required to excavate a 6.7 m ID tunnel), the minimum clear shaft diameters required to allow TBM launch and retrieval are 11 m and 9 m, respectively. Larger diameters may be required to accommodate the permanent structures or to suit the contractor's means and methods.

Key considerations in selecting construction methods include preventing groundwater drawdown and providing excavation support. Two methods of ground support will be used in the shafts. In the overburden soil layers, slurry wall panels, laid out to approximate a circle, will extend from the surface to the top of the competent rock. They will provide temporary support during construction as well as the permanent liner. In the rock, shafts will be excavated by drilling and blasting, and the rock face is supported using rock dowels and sprayed shotcrete.

Starter and tail tunnels will be required for TBM assembly, and to store equipment and muck cars, and will be excavated by drilling and blasting, with a horseshoe cross-section.

One and two-pass lining systems are both considered viable options for the SHCST. The final recommendation will depend on ground and groundwater conditions along the alignment, and the respective construction costs. The type of the tunnel final liner will be selected during the final design phase.

The anticipated ground conditions along the alignment necessitate the use of final lining, to meet the design criteria and ensure long-term stability, durability, and hydraulic performance. Viable options for SHCST are cast-in-place concrete (CIP) and precast concrete segments. Important considerations in lining selection include

• durability in the service environment without significant degradation during the tunnel's design life;

• constructability;

• life-cycle cost.

A quantitative approach, adopted by EPA and ASCE, will be used to assess the corrosion of the final concrete lining during the life-time operation of the tunnel. This approach enables estimates to be made of the rate of loss of liner material as a function of time, as well as its concrete properties and CSO characteristics.

The tunnel design recommendations are

• Define the geotechnical parameters for tunnel analysis and design.

• Perform groundwater infiltration and ground settlement analyses to quantify the risk of consolidation settlement due to dewatering.

• Analyze geotechnical data to support the selection of the tunnel lining system and type of TBM. Based on the available geotechnical information and construction cost estimate, both tunnel lining options, namely CIP concrete and precast concrete segmental lining, should be considered during final design.

Site plans were prepared to identify existing site conditions, areas for site access, staging and operations, work zone layouts and constraints, equipment and materials storage, utility protection and relocations, site drainage and grading, erosion and sedimentation controls, and electrical power requirements.

A conceptual, planning level, cost estimate, schedule and contract packaging was performed. Costs from similar, historical, projects were obtained and utilized to develop unit costs and extrapolated for the SHCST project. A detailed estimate was made of the construction cost of the main tunnel, and the TBM launch and retrieval shafts associated with Alignment F. Cost estimate options took into account different methods of tunnel excavation – double shield TBM with precast concrete segmental lining, and main beam TBM with CIP concrete lining. The cost estimate for the entire SHCST project is approximately $US500 M. The project construction duration is estimated at approximately 72 months (6 years).

The recommended contract packaging is to release six construction contracts: (1) preliminary utility relocation, (2) tunnel, (3) pumping station, and (4), (5), and (6) three separate consolidation conduit contracts. Contracts were grouped to align construction skill sets but allow for phased package release. The construction schedule will be coordinated so that the tunnel, pumping station, and consolidation conduits are constructed independently but finish at the same time.

The project goal is that odor complaints must not occur, so, the SHSCT control strategy is focused on minimizing odors from any of the shafts. Ventilation rates of approximately 2,300 to 2,400 m³/min have been estimated for both the upstream and downstream (main) shafts, and between 65 and 210 m³/min for the intermediate drop shafts.

Active fan driven odor control is recommended at the tunnel ends with passive systems for the six drop shafts, all using activated carbon for control/treatment. Footprint estimates indicate that the larger facilities at the tunnel ends will cover roughly 190 m² and the smaller systems, at the intermediate drop shafts, about 30 m².

Seven hydraulic drop shafts are used to transfer flow from the shallower conduits to the deep tunnel. A two-level screening process was used to assess each site's characteristics and recommend either a baffle-plunge or tangential vortex, based on cost effectiveness, hydraulic performance, and operation and maintenance considerations (Figure 5).

The tangential vortex drop was selected for all sites along the alignment (apart from the TBM retrieval site) due to its widely accepted use for deep rock CSO storage tunnels, performance history, and cost effectiveness when compared to baffle-plunge drop structures. The baffle-plunge type was selected for the deep tunnel TBM retrieval site because of the larger diameter shaft being constructed there. Once such a shaft is available, baffle-plunge is ideally suited because of its low surface area impact. Based on the drop shaft selections, potential operations criteria and maintenance requirements were developed for each site.

The consolidation conduits will be installed using a combination of micro-tunneling, guided boring, shallow rock tunneling, and open cut construction techniques. It is anticipated that three consolidation pipe branches along the alignment will be installed using micro-tunneling. This includes a 60 cm guided bore, and 105 and 120 cm micro-tunnel installations. When considering micro-tunneling, an effort has been made to locate conduits in soil, although there is potential for mixed-face micro-tunneling in areas of till.

The open cut method will be utilized for 75 and 70 cm consolidation pipes. The method causes more temporary disturbance to traffic, business and residences than micro-tunneling as such work is performed primarily within roadways. It might be the preferred installation method, however, due to the depth of the pipe, geotechnical conditions, and/or cost considerations. Open cut installations are typically shallower than micro-tunnels.

Geotechnical information suggests that it might be best to use an open face rock tunneling machine to construct the two other remaining consolidation pipes with 170 and 152 cm diameters. Tunnel diameter standardization will be considered during final design with respect to potential cost reductions.

The TPS is designed to dewater SHCST after storms so that the flow can be treated at HWPCF, where adequate treatment can be given prior to discharge to the Connecticut River. The TPS will be in HWPCF. The maximum design pumping rate will be 100 MLD, allowing the 235 ML SHCST to be dewatered within 55 hours (2.3 days). The proposed tunnel invert elevation at the TPS is – 52 m and the discharge elevation +2 m, giving a total, maximum static head of 54 m.

The recommended pumping equipment consists of four 35 MLD vertical non-clog centrifugal pumps, to provide a reliable pumping capacity of 100 MLD with one unit out of service. Variable frequency drives are recommended for the pumps as turn-down capability to approximately 15 to 20 MLD can be achieved on each unit.

The TPS will discharge directly to the new headworks being designed for HWPCF. The force main is expected to be 90 cm in diameter and the recommended connection point at the discharge end is at a new junction structure just upstream of the new influent pumping station. A surge tank will be provided on the discharge force main – on the TPS site – to minimize surges in the system.

Two pumping station configurations were presented as the final options – one a cavern and the other a circular shaft with a suction header pipe system (Figure 6). The two are comparable in cost but the cavern pumping station has some advantages not related to cost, mainly concerning the maintenance of crane lifts. If the circular shaft pumping station layout is fitted with a bridge crane at the lower level, however, the configurations become essentially the same from a maintenance perspective. Following other design modifications to meet safety and other requirements, a comparative assessment of their capital costs showed the cost of the cavern pumping station to be approximately $US7 M lower than that of the circular shaft design, so the cavern was recommended.

A new, 9,800 kW overhead power service will be required for the TBM. This will be converted to a permanent feed for the TPS, once that is complete and operational. Power requirements for the TPS are estimated at 3,055 kW.

Screening and grit capture will be accomplished in a separate, 11 m diameter, dedicated shaft, which will be used to launch the TBM and converted after drilling. Bar screens will protect the TPS pumps from solids and debris, which could either clog or damage them. A rake lowered by crane will push or pull the screenings up from the shaft, while grit and heavier debris will be removed by a clamshell bucket. The screenings shaft will be used for tunnel and TPS construction, allowing them to proceed in parallel.

The TPS and grit/screening facility will be roughly 45 m apart and connected by a 120 cm diameter suction header. An at-grade building over the sub-surface pumping station will house support facilities and enable access to the pumping station. Personnel access/egress will be by elevator, with a separate stair tower for emergencies. The grit/screening facility will also be in a building for odor containment and to make it more attractive visually.

This paper presents the design of a deep rock conveyance and storage tunnel, drop shafts, consolidation conduits, and a pumping station in Hartford, CT. The geological settings and sub-surface investigation program are discussed, and the general aspects of the preferred alignment selection are described. Relevant alternatives for the drop shafts and the pumping station are explained, and recommended options are presented.